Salinity Changes and Anoxia Resulting from Enhanced Runoff During the Late Permian Global Warming and Mass Extinction Event

Home , Acritarch

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

1 Salinity changes and anoxia resulting from enhanced runoff 2 during the late Permian global warming and mass extinction 3 event 4

5 Elsbeth E. van Soelen1, Richard .J Twitchett2, Wolfram M. Kürschner3 6 7 1University of Oslo, Departments of Geosciences, P.O box 1047 Blindern 0316 Oslo, Norway 8 2Natural History Museum, Earth Sciences Department, London, SW7 5BD, UK 9 3Department of Geosciences and Centre for Earth Evolution and Dynamics (CEED), University of Oslo, Oslo, 10 Norway 11 12 Correspondence to: Elsbeth E. van Soelen ([email protected])

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

14 Abstract. The Late Permian biotic crisis had a major impact on marine and terrestrial environments. Rising CO2 15 levels following Siberian Trap volcanic activity were likely responsible for expanding marine anoxia and 16 elevated water temperatures. This study focusses on one of the stratigraphically most expanded Permian- 17 Triassic records known, from Jameson land, east Greenland. High resolution sampling allows for a detailed 18 reconstruction of the changing environmental conditions during the extinction event and the development of 19 anoxic water conditions. Since very little is known about how salinity was affected during the extinction event, 20 we especially focus on the aquatic palynomorphs and infer changes in salinity from changes in the assemblage 21 and morphology. The extinction event, here defined by a peak in spore/pollen, indicating disturbance and 22 vegetation destruction in the terrestrial environment, postdates a negative excursion in the total organic carbon, 23 but predates the development of anoxia in the basin. Based on the newest estimations for sedimentation rates, 24 the marine and terrestrial ecosystem collapse took between 1.6 to 8 kyrs, a much shorter interval than previously 25 estimated. The palynofacies and palynomorph records show that the environmental changes can be explained by 26 enhanced runoff, increased primary productivity and water column stratification. A lowering in salinity is 27 supported by changes in the acritarch morphology. The length of the processes of the acritarchs becomes shorter 28 during the extinction event and we propose that these changes are evidence for a reduction in salinity in the 29 shallow marine setting of the study site. This inference is supported by changes in acritarch distribution, which 30 suggest a change in palaeoenvironment from open marine conditions before the start of the extinction event to 31 more near-shore conditions during and after the crisis. In a period of sea-level rise, such a reduction in salinity 32 can only be explained by increased runoff. High amounts of both terrestrial and marine organic fragments in the 33 first anoxic layers suggest that high runoff, increased nutrient availability, possibly in combination with soil 34 erosion, are responsible for the development of anoxia in the basin. Enhanced runoff could result from changes 35 in the hydrological cycle during the late Permian extinction event, which is a likely consequence of global 36 warming. In addition, vegetation destruction and soil erosion may also have resulted in enhanced runoff. 37 Salinity stratification could potentially explain the development of anoxia in other shallow marine sites. The 38 input of fresh water and related changes in coastal salinity could also have implications for the interpretation of 39 oxygen isotope records and sea water temperature reconstructions in some sites.

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

41 1. Introduction

42 The late Permian extinction event was the most severe global crisis of the Phanerozoic in terms of both 43 taxonomic loss and ecological impact (e.g. McGhee et al., 2012). The current consensus is that the extinction

44 was likely due to global warming and associated environmental changes caused by CO2 emissions from Siberian 45 Trap volcanic activity, because of the close timing between the volcanic activity and the extinction event (e.g., 46 Burgess et al., 2017; Burgess and Bowring, 2015). Most studies on the late Permian extinction have inferred that 47 expanding marine anoxia (e.g. Wignall and Hallam, 1992) is a key biotic factor causing marine extinction and 48 ecosystem collapse. There are different theories to explain the spreading of anoxia, which affected both deep 49 and shallow sites (e.g. Bond and Wignall, 2010; Isozaki, 1997; Wignall and Twitchett, 1996). A weakened 50 temperature gradient between equator and pole would have slowed ocean circulation and may have facilitated 51 expansion of the oxygen minimum zone (Hotinski et al., 2001). Increased weathering and detrital input (Algeo 52 and Twitchett, 2010), and soil erosion (Sephton et al., 2005) led to enhanced terrestrial matter input in marine 53 sections, and may also have contributed to eutrophication and in stratified waters led to hypoxia or anoxia 54 (Sephton et al., 2005). Other factors that are thought to play important roles in the extinction are elevated water 55 temperatures (e.g. Sun et al., 2012), and ocean acidification (Clapham and Payne, 2011). One important 56 environmental parameter that has received relatively little attention is salinity, even though low salinity ocean 57 conditions were once considered to be a leading cause of the marine extinction (Fischer, 1964; Stevens, 1977). 58 Furthermore, potential impacts of changes in salinity, which might be expected from enhanced discharge of 59 freshwater into shelf seas (Winguth and Winguth, 2012), have been largely ignored.

60 Microfossils of marine algae are excellent recorders of environmental change in the water column. Of 61 these, the organic-walled cysts of dinoflagellates have proven especially useful in palaeo-environmental studies 62 (e.g. Ellegaard, 2000; Mertens et al., 2009, 2012, Mudie et al., 2001, 2002; Sluijs and Brinkhuis, 2009; Vernal et 63 al., 2000). While dinocysts are absent or sparse in Palaeozoic deposits, acritarchs are commonly recorded (e.g. 64 Tappan and Loeblich, 1973), even during the late Permian when acritarch diversity was declining (Lei et al., 65 2013b). Acritarchs are a group of microfossils of organic composition and unknown affinity (Evitt, 1963). Many 66 acritarchs are, however, found exclusively in marine rocks, and some are thought to be precursors of modern 67 dinoflagellate cysts (e.g. Servais et al., 2004). Several studies in East Greenland have reported fluctuating 68 abundances of acritarchs during the Late Permian and Early Triassic (e.g. Balme, 1979; Piasecki, 1984; 69 Stemmerik et al., 2001), but very few late Permian studies have documented the relative abundance of the 70 different genera of aquatic palynomorphs (for example, Shen et al., 2013). Similar to dinocysts, acritarch 71 morphology has been linked to environmental conditions, and acritarchs with longer processes are generally 72 more abundant in more open marine settings (Lei et al., 2012; Stricanne et al., 2004). Both dinocyst and 73 acritarch studies show that salinity is an important, albeit not the only, factor that influences cyst morphology. It 74 is thought that the longer processes in higher salinities stimulate clustering with other cysts or particles in the 75 water column, and thus enhance sinking to the seafloor (Mertens et al., 2009).

76 The rock record of Jameson Land, East Greenland, has provided key insights into marine environmental 77 changes during the late Permian extinction event. Previous work on this section by Twitchett et al. (2001) 78 showed that collapse of marine and terrestrial ecosystems was synchronous and took between 10 to 60 kyr. 79 Palynological work by Looy et al. (2001) showed that the terrestrial ecosystem collapse is characterized by a

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

80 distinct rise in spores, indicative of the loss of woodland and increase in disturbance taxa. The onset of anoxic 81 conditions is not as abrupt at this location as seen in some other sections, but instead Wignall and Twitchett, 82 (2002) found unusual alternating patterns of bioturbated and laminated siltstones in the top of the Schuchert Dal 83 Formation and lowest of the Wordie Creek Formation (Fig. 1). In a recent study, Mettam et al., (2017) showed 84 that rapidly fluctuating redox conditions occurred during the late Permian extinction, while sedimentological 85 observations imply that sea level was rising in this period. Estimations of sedimentation rates indicate that each 86 bioturbated/laminated rock interval, which are between 2-10 cm thick, have been deposited in a period of 50- 87 1000 years (Mettam et al., 2017). To study the environmental conditions associated with the deposition of these 88 sediments we look in detail at a short (9 m) interval covering the late Permian extinction and the Formation 89 boundary between the Schuchert Dal and Wordie Creek Formation. Palaeoenvironment is studied by looking at 90 variations in the palynofacies (organic particles) and the acritarch assemblage. In addition we present a record of 91 morphological changes within the acritarch Micrhystridium. Runoff and erosion, salinity, sea-level fluctuations 92 and temperature rise are discussed as possible reasons for the changing environmental conditions.

94 2. Geological setting and stratigraphy

95 Samples were collected at the Fiskegrav location of Stemmerik et al. (2001), an outcrop in East Greenland 96 (N71°32’01.6ˮ, W024°20’03.0ˮ), in a small stream section on the east side of Schuchert Dal (Fig. 1). The 97 section was deposited within a relatively narrow, north-south oriented basin (Stemmerik et al., 2001; Wignall 98 and Twitchett, 2002). Active rifting and rapid subsidence has resulted in the deposition of one of the most 99 expanded P-Tr sections. At the Fiskegrav location there are no obvious breaks in sedimentation, even though 100 large erosive channels exist in P-Tr sections in more northern locations (Twitchett et al., 2001; Wignall and 101 Twitchett, 2002).

102 The Fiskegrav section shows a transition from the Schuchert Dal Formation into the Wordie Creek 103 Formation. The Permian/Triassic boundary, defined by the first occurrence of the conodont Hindeodus parvus 104 (Yin, 1996), is located at 23.5 m above the base of the Wordie Creek Formation (Twitchett et al., 2001). This 105 study focusses on a ca. 9 m section interval including the late Permian extinction and the formation boundary 106 (Fig. 1). The upper part of the Schuchert Dal Formation consists of blocky (bioturbated), micaceous, greenish 107 mudstones (muddy siltstones), whereas the lower Wordie Creek Formation consists mainly of laminated, dark 108 grey mudstones (clay-siltstones) that often contain framboidal pyrite (Twitchett et al., 2001; Wignall and 109 Twitchett, 2002). The extinction event occurs in the upper metres of the Schuchert Dal Formation, and the start 110 of the biotic crisis is synchronous in marine and terrestrial ecosystems (Looy et al., 2001; Twitchett et al., 2001). 111 For more details on biostratigraphy see Mettam et al., (2017); Stemmerik et al., (2001) and Twitchett et al., 112 (2001).

113

114 3. Material and Method

115 3.1 Material

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

116 A total of 36 samples were collected from the top 4.5m of the Schuchert Dal Formation and the lower 4.2m of 117 the Wordie Creek Formation. Sample resolution is highest (10cm intervals) in the 2.5m interval covering the 118 formation boundary and the extinction event. Depths are given in m, relative to the formation boundary between 119 Schuchert Dal and Wordie Creek Fm. Some of the samples used in this study were also analysed by Mettam et 120 al. (2017) during their study of changes in redox conditions at the formation boundary.

121

122 3.2 palynofacies and palynological analysis

123 Samples (of ca.5 g) were crushed, and a tablet with a known amount of lycopodium spores was added 124 to allow quantification of organic particles and palynomorphs. Samples were then treated with hydrochloric acid 125 and hydrofluoric acid to remove carbonates and silicates, and subsequently sieved through 7 µm sieves. Heavy 126 liquid separation was used to remove heavy minerals like pyrite. The residue, containing palynomorphs and 127 other organic particles, was mounted onto microscope slides. Material was counted using a Leitz Diaplan 128 microscope and an AxioCam ERc 5s camera attached to a computer with Zen microscope software (Zen 2 lite, 129 2011). A minimum of 300 particles per sample were counted for palynofacies analyses. Aquatic palynomorphs 130 were counted to reach 100 specimens when abundant and at least 30 specimens when abundance was low. Still, 131 due to low abundances during the biotic crisis, in some samples counts were below 30 (these 6 samples are 132 indicated in Fig. 4).

133

134 3.3 Acritarch measurements

135 The genus Micrhystridium includes many species, of which ca. 20 are known from the Permian- 136 Triassic interval, which differ in their body size, number of processes and process length (Lei et al., 2013b; 137 Sarjeant, 1970). Only those with a spherical central body and many (>10) simple processes with closed tips, 138 were included in this study (i.e. the M. breve Group, according to Lei et al., 2013b). Thus, amongst others, M. 139 pentagonale was excluded. Specimens were excluded if they were too damaged (folded or broken) to allow 140 measurements. To counteract the effects of compression (which can make the body look longer), body size was 141 calculated from the average of two linear measurements perpendicular to each other. For each specimen three 142 processes were measured (all within one focal plane). The aim was to measure 20-30 specimens per sample, but 143 acritarch concentrations were low during the biotic crisis interval, and therefore 5 samples had lower counts 144 (indicated in Fig. 6).

145

146 4. Results

147 4.1 Palynofacies

148 The most abundant organic particles are phytoclasts and amorphous organic matter (AOM) which together 149 comprise 45 to 90% of the palynofacies (Fig. 2). Phytoclasts are dominant in the bioturbated samples in the

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

150 lower part of the section, while AOM dominates the palynofacies in the laminated intervals. Other terrestrially 151 derived organic (wood fragments and charcoal) particles make up a small portion (<20%) of the palynofacies 152 and are not influenced by the alternating anoxic/oxic depositional conditions (Fig. 2). Palynomorphs were 153 divided into three groups: pollen, spores and aquatic palynomorphs. Pollen abundance relative to other 154 palynofacies is higher in the upper 0.5 m of the Schuchert Dal Formation and lowest metre of the Wordie Creek 155 Formation (5-20%), which is mainly caused by the high number of pollen fragments (sacci) during and 156 following the extinction event (Fig. 2). Spores have highest abundance in the upper metre of the Schuchert Dal 157 Formation (up to 35%), where also the spore/pollen ratio is highest. Aquatic palynomorphs (acritarchs and other 158 algal remains like Tasmanites and Cymatiosphaera) make up a very small fraction of the palynofacies in the 159 Schuchert Dal Formation (<2%), but become relatively more abundant in the Wordie Creek Formation (up to 160 14%).

161

162 4.2 Aquatic palynomorphs

163 Although Permian acritarch diversity is considered to be low, an overview of Permian phytoplankton by Lei et 164 al. (2013b) showed a richness of about 20-30 genera during most of the Permian, with acritarchs belonging to 165 the genera Micrhystridium and Veryhachium being the most common (Lei et al., 2013b, 2013a; Sarjeant and 166 Stancliffe, 1994). Since cysts are generally small in size it is difficult to study them under light microscope and 167 identification at species level can be difficult (Sarjeant, 1970). Several revisions and simplifications have been 168 proposed for the genera (Lei et al., 2013a; Sarjeant and Stancliffe, 1994). We follow the simple classification 169 scheme proposed by Lei et al. (2013a), in which different species of acritarchs are grouped together, based on 170 geometrical shape of the acritarchs. Photos of selected aquatic palynomorphs are shown in the plate of Fig. 3.

171 Concentrations of aquatic palynomorphs were on average ca.4 times higher in the laminated rocks 172 compared to the bioturbated rocks (Fig. 4). Acritarch abundance also declines during the extinction event. The 173 majority (25 - 100%) of the acritarchs belong to the Micrhystridium breve Group (Fig. 4). This group includes 174 all acritarchs with a spherical central body and numerous processes. The Veryhachium laidii Group (acritarchs 175 with a rectangular central body shape), the V. trispinosum Group (triangular shape), V. cylindricum Group 176 (ellipsoidal shape) and M. pentagonale Group (pentagonal or hexagonal shape) were mostly found in 177 bioturbated intervals of the upper part of the Schuchert Dal Formation, especially just before and during the 178 extinction interval (Fig. 4). The leiospheres include all aquatic palynomorphs with a spherical shape and either a 179 smooth surface or with small ornamentation, but no processes. The highest abundances (up to 70%) of 180 leiospheres are found in the laminated sediments of the Wordie Creek Formation (Fig. 4). Cymatiosphaera are 181 rare in the record and found in both the oxidized siltstones of the Schuchert Dal Formation, but also in some 182 intervals in the laminated siltstones. Tasmanites were found only in one sample, also during the marine 183 extinction phase. The apparent relatively higher abundance of Cymatiosphaera and Tasmanites in the top of the 184 Schuchert Dal Formation is a consequence low total count of the aquatic palynomorphs during the extinction 185 event. The most diverse samples were found in the bioturbated rocks of the upper metre of the Schuchert Dal 186 Formation, an interval that partly overlaps with the extinction interval.

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

187 The Micrhystridium Group has a diverse morphology, and shows variation in vesicle size and number 188 of processes. Most of the measured specimens are very similar to the species M. breve as described by Jansonius 189 (1962) in both the number and shape of the processes. However, the size of the central body and the length of 190 the processes of many of the measured acritarchs fall outside of the defined size-range of M. breve. The 191 distributions of both body size and process length are unimodal (Fig. 5), which strongly suggests that the 192 measured specimens all belong to a single species of Micrhystridium. Body size is variable and in the two-metre 193 interval spanning the formation boundary, acritarchs in the laminated rocks are on average larger compared to 194 those in the bioturbated rocks (Fig. 6). Process length decreases gradually from the start of the extinction event 195 until its end, and then remains shorter through to the top of the studied section. Process length not only changes 196 in absolute terms, but also relatively to the size of the body.

197

198 5. Discussion

199 5.1 The extinction event and onset of anoxia

200 Consistent with the results of Looy et al. (2001), we find a strong increase in spores, relative to pollen, 201 in the upper Schuchert Dal Formation (Fig. 2). Since this peak in spores coincides with the interval of marine 202 ecosystem collapse defined by Twitchett et al. (2001), we define the period with highest spore/pollen ratios as 203 the extinction interval. Twitchett et al. (2001) found large changes in marine plankton abundance in their study 204 of the Fiskegrav section, with a bloom just before the disappearance of trace fossils, followed by almost 205 complete absence of acritarchs immediately after the collapse of the marine ecosystem. This phytoplankton 206 bloom is not obvious in our higher resolution data, and may simply be an artefact of the relatively low- 207 resolution sampling of Twitchett et al., (2001), but the number of acritarchs is indeed greatly reduced during the 208 biotic crisis (Fig. 4). The peak in spore/pollen and decrease in acritarch abundance occurs in an interval of ca. 209 0.8m, which is comparable to the interval found by Twitchett et al., (2001) for marine and terrestrial ecosystem 210 collapse. Twitchett et al., (2001) estimated the collapse to have taken between 10 and 60 kyrs, however, recently 211 improved estimations of sedimentation rates in the Fiskegrav section (Mettam et al 2017) now give an estimated 212 duration of between 1.6 to 8 kyrs for the marine and terrestrial ecosystem collapse.

213 During the start of the biotic crisis rocks remain bioturbated but within the upper 0.5 m of the 214 Schuchert Dal Formation there are intervals with laminated rocks (Fig. 2), indicative for anoxia (Twitchett et al., 215 2001). AOM, which is structureless organic matter, is dominant in the laminated rocks. AOM can derive from 216 degraded macrophyte, phytoplankton or higher plant tissue, or it may also be bacterially derived (Tyson, 1995). 217 Anoxic conditions can stimulate the production of marine AOM (Pacton et al., 2011), but also result in better 218 preservation after sedimentation (Tyson, 1995). The relative proportions of AOM versus phytoclasts thus 219 indicate the amount of oxygenation of the depositional environment (Tyson, 1995). A recent, high-resolution 220 study of the δ13C of total organic carbon (TOC) by Mettam et al., (2017) shows a negative shift of 5-6‰ that 221 begins 1.80m below the top of the Schuchert Dal Formation at Fiskegrav, (Fig. 2). The onset of this negative 222 excursion coincides closely with the last occurrence of Permian macroinvertebrate fossils (Mettam et al., 2017), 223 but pre-dates the collapse of marine and terrestrial ecosystems defined by disappearance of bioturbation.

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

224 Mettam et al. (2017) record a small, but consistent, offset in δ13C values between the laminated/bioturbated 225 intervals (Fig. 2). This offset can be explained by the differences in organic matter source, which is mostly 226 phytoclasts (terrestrial) in the bioturbated intervals while AOM (marine and terrestrial) dominates in the 227 laminated intervals. However, the main shift is not associated with major changes in the palynofacies, and 228 occurs in a part of the section where the palynofacies mainly consist of terrestrially derived organic matter (Fig. 229 2). This negative shift in carbon isotopes associated with the late Permian extinction event is recognized 230 worldwide (e.g. Korte and Kozur, 2010) and usually precedes, the extinction event (e.g. Burgess and Bowring, 231 2015; Korte and Kozur, 2010). It has been interpreted as resulting from atmospheric changes in δ13C values, 232 which can be directly, or indirectly, related to Siberian trap volcanism (e.g. Cui and Kump, 2015; Svensen et al., 233 2009).

234 Mettam et al. (2017) studied redox conditions in the same section and showed that conditions became 235 potentially anoxic, and ferruginous (Fe2+- rich) in the laminated intervals near the formation boundary. Within 236 the lowermost Wordie Creek Formation, from 0.6-0.7 m above the formation boundary, anoxic conditions are 237 confirmed by Fe speciation (Mettam et al., 2017). Since the extinction started before conditions became 238 ferruginous, spreading anoxia cannot be the main cause for the marine biotic crisis in the Fiskegrav section. 239 Grasby et al. (2015) distinguished three phases of extinction in the shallow marine sequence of Festningen 240 (Svalbard): first is the loss of carbonate macrofauna, followed by loss of siliceous sponges, and finally a loss of 241 all trace fossils. Thus, similar to what is observed in East Greenland, the first laminated sediments do not 242 represent the start of the biotic crisis. This is unsurprising as expanding marine hypoxia is a predicted

243 consequence of global warming, which was probably caused by elevated CO2 flux from Siberian Trap 244 volcanism (Benton and Twitchett, 2003; Grard et al., 2005). Therefore, extinctions that occurred prior to the 245 appearance of widespread anoxia were more likely due to the direct, short-term effects of volcanic activity (e.g.

246 metal toxicity), or the more immediate effects of CO2 rise or temperature increase. The laminated rocks are 247 associated with higher amount of pollen fragments (Fig. 2), which could be an indication for elevated terrestrial 248 organic matter input (e.g. soil erosion), while the higher amount of AOM can be partly explained by terrestrial 249 organic matter input, and partly by increased marine productivity and better preservation of organic matter due 250 to low oxygen conditions. Increased weathering and run-off are expected consequences of atmospheric and 251 hydrological changes associated with global warming. Algeo and Twitchett, (2010) demonstrated that sediment 252 accumulation rates greatly increased in shallow shelf seas in the latest Permian and earliest Triassic, consistent 253 with enhanced weathering and run-off, and Sephton et al. (2005) found evidence for large-scale soil erosion.

254

255 5.2 Sea-level and salinity changes

256 If the laminated rocks are indeed a consequence of enhanced runoff, this is expected to affect also 257 marine environmental conditions, such as a decrease in (surface) salinity. At the same time, sea-level 258 fluctuations at and near the boundary need to be considered, since these would also affect marine environmental 259 conditions and could potentially affect salinity. The Fiskegrav section has well preserved aquatic palynomorphs, 260 and their diversity, distribution and morphology can provide valuable information on marine environmental 261 changes. Palaeozoic studies have shown that acritarch diversity is generally higher in deeper, more distal

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

262 settings (e.g. Lei et al., 2012; Stricanne et al., 2004). In addition, different studies have shown that Veryhachium 263 favoured more open marine settings, while Micrhystridium favoured near-shore environments (e.g. Wall 1965, 264 Lei et al., 2012). It is possible that some of the leiospheres in the Fiskegrav section belong to the genus 265 Leiosphaeridia, as the dimensions are comparable to L. microgranifera and L. minutissima that were found in 266 Chinese Permian-Triassic sections (Lei et al., 2012), but the small ornamentations on both of those species is 267 different (Fig. 3). In these Chinese Permian sections the Leiosphaeridia are associated with deeper/more open 268 marine waters (Lei et al., 2012). On the other hand, studies of early Palaeozoic acritarchs show that, in general, 269 leiospheres, or sphaeromorphs, are most frequent in proximal environments (Li et al., 2004; Stricanne et al., 270 2004). The leiosphere-group might be polyphyletic and some species might have prasinophyte affinities 271 (Colbath and Grenfell, 1995), which complicates their use for environmental reconstructions. Whatever the 272 biological origin of the leiosphere is, in this study site they apparently favoured the palaeoenvironmental 273 conditions associated with the deposition of the laminated rocks, starting during the extinction event. 274 In the Fiskegrav section, the decline in diversity from the upper Schuchert Dal Formation to the Wordie 275 Creek formation, together with the change in assemblage from Veryhachium/Micrhystridium to 276 Micrhystridium/leiosphere dominance could thus be interpreted as a change from a distal to a more proximal 277 setting, which would suggest a marine regression. Although this was once considered as a possible cause of the 278 late Permian extinction event (Hallam and Cohen, 1989), the current consensus is that eustatic sea-level fall 279 occurred prior to collapse of marine ecosystems and the extinction, and the subsequent transition from the 280 Permian to the Triassic happened during sea-level rise (Wignall and Hallam, 1992). In some East Greenland 281 locations, apparently missing biozones and local erosive conglomeratic channels have been proposed as 282 evidence for sea-level fall near the boundary between the Schuchert Dal and Wordie Creek formations (Surlyk 283 et al., 1984), but Wignall and Twitchett (2002) demonstrated that in central Jameson Land the sedimentological 284 changes are consistent with sea-level rise, and that the erosive, conglomeratic channels occur at multiple 285 stratigraphic levels and postdate the extinction and Permian/Triassic boundary. 286 Thus, during the extinction event the acritarch assemblage changes to a more typical near-shore 287 assemblage, despite the ongoing sea-level rise. Instead of a marine regression, the observed shift in acritarch 288 assemblages could be explained by an increase in runoff which affects the near-shore environment by lowering 289 surface water salinity and delivering nutrients, both of which are found to affect the distribution of modern 290 phytoplankton (e.g. Bouimetarhan et al., 2009; Devillers and de Vernal, 2000; Pospelova et al., 2004; Zonneveld 291 et al., 2013). Dinocyst morphology has also been linked to environmental conditions (e.g., Mertens et al., 2009, 292 2011), with salinity identified as an important factor influencing process length (Ellegaard et al., 2002; Mertens 293 et al., 2011). Similarly, studies of acritarch process length and palaeoenvironment show that species and 294 individuals with longer processes are generally found in more offshore locations, while in inshore settings 295 acritarchs with shorter processes are more abundant (Lei et al., 2013a; Servais et al., 2004; Stricanne et al., 296 2004). The average process length of specimens belonging to the Micrhystridium breve Group, show a 297 decreasing trend during the extinction interval, with higher values before the extinction event, and lower values 298 after the event (Fig. 6), which could thus be interpreted as a lowering in salinity during the extinction event. 299 Within the 2m interval including the extinction interval (-1 – +1m) there is also a consistent offset in body size 300 between samples from bioturbated and laminated rocks. The body size of some dinocysts has been found to vary 301 with temperature and salinity (e.g. Ellegaard et al., 2002; Mertens et al., 2012), however, the relation has not

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

302 been well studied. Possibly low oxygen or salinity conditions or increased nutrient availability influenced 303 acritarch body size in this interval. Runoff, lower salinity, and salinity stratification together can explain the 304 observed changes in palynofacies and acritarch records. The changed environmental conditions continued 305 through the aftermath of the biotic crisis, as shown by persistently low process lengths to the end of the study 306 interval (Fig. 6). Meanwhile, the ongoing marine transgression likely resulted in the shoreward migration of 307 oxygen deficient waters which explain the development of permanently anoxic conditions higher up in the 308 Wordie Creek Formation (over 0.6-0.7 m from the Fm boundary) (Mettam et al., 2017).

309 The development of widespread anoxia in the late Permian is usually explained by an expansion of 310 anoxic deep waters onto the shelf (e.g. Grasby and Beauchamp, 2009). However, increased runoff and the 311 development of estuarine circulation provide an alternative explanation for the development of anoxia in some 312 shallow marine environments. Furthermore, low salinity is known to negatively impact biomass and size of 313 modern marine invertebrates (e.g. Westerbom et al., 2002), and, along with higher temperatures and hypoxia, 314 might also be a contributing factor in the size reduction of marine organisms recorded in the aftermath of the 315 late Permian extinction event, such as observed for the bivalve Claraia the Permian-Triassic section of East- 316 Greenland (Twitchett, 2007). Low salinity conditions have previously been invoked to explain the dominance of 317 the brachiopod lingulid in post-extinction ecosystems (Zonneveld et al., 2007), because they are able to tolerate 318 a wide range of salinities as well as hypoxia (e.g. Hammen and Lum, 1977). Lingulid fossils are found in many 319 Permian-Triassic (shallow) marine sections (e.g. Peng et al., 2007) and the lowering of salinities in shallow 320 marine settings and spreading anoxia provided ecological space for the lingulids (Peng et al., 2007; Rodland and 321 Bottjer, 2001). It has been suggested that rising sea water temperatures in the Tethys, resulted in strengthened 322 monsoonal activity (e.g. Winguth and Winguth, 2012) and enhanced precipitation and runoff in the areas 323 surrounding the Tethys (e.g. Parrish et al., 1982). Whether this would have had an effect on the study site is 324 questionable, because of its distance to the Tethys and relatively sheltered, inland position. Possibly river 325 systems existed which brought freshwater to the Greenland-Norway basin from the South. It is also possible that 326 rainfall increased locally. Mettam et al., (2017) speculated that changes in circulation patterns may have resulted 327 in onshore winds bringing moist air to the study site. In addition vegetation destruction and soil erosion may 328 also have contributed to enhanced runoff.

329

330 5.3 Sea water temperature

331 Besides salinity, increasing water temperatures can also affect process length in dinocysts (e.g. Mertens 332 et al., 2009, 2012) but this relationship is not always evident (e.g. Ellegaard et al., 2002). Sea water temperature 333 reconstructions for the late Permian and Early Triassic exist mainly from carbonate rich sections originating in 334 the Tethys Ocean (e.g. Joachimski et al., 2012; Schobben et al., 2014; Sun et al., 2012). These reconstructions 335 indicate increasing temperatures during the extinction event, and maximums temperatures are reached in the 336 Early Triassic (e.g. Joachimski et al., 2012; Kearsey et al., 2009; Schobben et al., 2014; Sun et al., 2012). Since 337 no water temperature reconstructions exist for the Greenland-Norway basin, at this time, it is not possible to 338 exclude the effects of rising water temperatures on the acritarch morphology. Increasing water temperatures 339 might be partly responsible for decreasing process length in acritarchs, however, a rising sea-level in

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

340 combination with rising temperatures are not able to explain the changes in the acritarch assemblages, which 341 indicate a shift from more open marine to near-shore conditions. Increased runoff cannot only explain the 342 reconstructed changes in palynofacies and acritarch records, but in addition, a reduction in salinity also has 343 significant implications for palaeotemperature reconstructions calculated from the oxygen isotopes of carbonates. 344 Most Permian-Triassic temperature records are from the Tethys or surrounding areas, where monsoon activity 345 has been proposed to increase (Winguth and Winguth, 2012) and thus salinity is expected to change as well. 346 Permian-Triassic oxygen isotope data for palaeotemperature estimates have been derived from brachiopod shells 347 (Kearsey et al., 2009) or conodont apatite (Joachimski et al., 2012; Schobben et al., 2014), with all studies 348 concluding that temperature rose through the extinction event and that Early Triassic seas were 'lethally hot' 349 (Sun et al., 2012). Oxygen isotope values partially reflect seawater salinity, although this is generally ignored in 350 palaeotemperature studies, because there is currently no independent proxy of salinity change, and so it is 351 assumed that salinity remained constant. If salinity was reduced in shelf seas due to hydrological changes (e.g. 352 enhanced precipitation and run-off), as it appears for East Greenland at least, then all of these studies may have 353 significantly over-estimated Early Triassic temperatures. The actual temperature rise associated with the 354 extinction event would have been lower and below 'lethal' levels, which is more consistent with fossil evidence 355 that the oceans were not completely devoid of life.

356

357 6. Conclusions

358 The highest resolution palynological sampling yet attempted for the Fiskegrav location of central Jameson Land 359 has shown significant, millennial-scale changes in marine and terrestrial environments. The biotic crisis, 360 defined by a peak in the spore/pollen ratio which indicates widespread disruption and collapse of the terrestrial 361 flora, postdates the onset of the carbon-isotope excursion. The spore/pollen peak spans 0.8 m of section, which, 362 on current age estimates, gives an estimation of 1600 to 8000 years for the ecosystem collapse. This is much 363 shorter than previous estimations. The onset of the biotic crisis does not appear to coincide with any obvious 364 changes in lithology or sedimentary facies, and deposition of the shallow marine shelf sediments appears to have 365 taken place under oxygenated waters. During the latter stages of the biotic crisis, laminated beds begin to appear, 366 indicating periodic anoxic deposition. These laminated rocks contain palynological evidence for a more near- 367 shore setting, which contradicts the widely accepted evidence that globally sea-level was actually rising during 368 the extinction event. A preferred alternative explanation, consistent with palynological assemblages, is that 369 enhanced runoff, water column stratification and enhanced primary productivity led to the development of 370 anoxic bottom waters. Enhanced runoff would lead to elevated freshwater flux into shallow shelf seas, which 371 would have led to the development of a freshwater wedge, stratification and a reduction in salinity. Acritarch 372 process length is considered here to be a proxy for seawater salinity, and supports these inferred changes. 373 Increased runoff did not immediately lead to anoxia, possibly because precipitation gradually increased during 374 the extinction event until the freshwater flow was large enough to cause water column stratification. Increasing 375 sea water temperatures may have also affected acritarch morphology, and partially be responsible for the 376 reduced process length. However, rising temperatures alone cannot explain the reconstructed changes in the 377 palynofacies and acritarch records. In fact, sea water temperature reconstructions may overestimate the rise in

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

378 water temperatures during the extinction event, since possible changes in salinity are not included in the 379 calculations.

380

381 7. Data availability. The data reported in this paper is available in tables in the supplementary files.

382

383 Author contributions. WMK designed the study. RJT provided material. EEvS analysed the samples for 384 palynofacies and palynomorph content and performed measurements on acritarchs. EEvS wrote the paper with 385 input from WMK and RJT.

386 387 Competing interests. The authors declare that they have no conflict of interest. 388

389 Acknowledgements

390 This work was funded by the Norwegian Research Council, project 234005, The Permian/Triassic evolution of 391 the Timan-Pechora and Barents Sea Basins. We thank B.L. Berg and M.S. Naoroz for technical support. 392 Samples were collected by RJT, who thanks G. Cuny and the Danish National Research Foundation for logistic 393 and financial support.

394

395 References

396 Algeo, T. J. and Twitchett, R. J.: Anomalous Early Triassic sediment fluxes due to elevated weathering rates and 397 their biological consequences, Geology, 38(11), 1023–1026, doi:10.1130/G31203.1, 2010.

398 Algeo, T. J., Heckel, P. H., Maynard, J. B., Blakey, R. and Rowe, H.: Modern and ancient epeiric seas and the 399 super-estuarine circulation model of marine anoxia, Dynamics of Epeiric seas: sedimentological, paleontological 400 and geochemical perspectives. Geological Association Canada Special Publication, 7–38, 2008.

401 Balme, B. E.: Palynology of Permian-Triassic boundary beds at Kap Stosch, East Greenland, Nyt nordisk forlag, 402 200(6), 1979.

403 Benton, M. J. and Twitchett, R. J.: How to kill (almost) all life: the end-Permian extinction event, Trends in 404 Ecology & Evolution, 18(7), 358–365, doi:10.1016/S0169-5347(03)00093-4, 2003.

405 Bond, D. P. G. and Wignall, P. B.: Pyrite framboid study of marine Permian–Triassic boundary sections: A 406 complex anoxic event and its relationship to contemporaneous mass extinction, GSA Bulletin, 122(7–8), 1265– 407 1279, doi:10.1130/B30042.1, 2010.

408 Bouimetarhan, I., Marret, F., Dupont, L. and Zonneveld, K.: Dinoflagellate cyst distribution in marine surface 409 sediments off West Africa (17–6°N) in relation to sea-surface conditions, freshwater input and seasonal coastal 410 upwelling, Mar. Micropaleontol., 71(3–4), 113–130, doi:10.1016/j.marmicro.2009.02.001, 2009.

411 Burgess, S. D. and Bowring, S. A.: High-precision geochronology confirms voluminous magmatism before, 412 during, and after Earth’s most severe extinction, Science Advances, 1(7), e1500470, 413 doi:10.1126/sciadv.1500470, 2015.

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

414 Burgess, S. D., Muirhead, J. D. and Bowring, S. A.: Initial pulse of Siberian Traps sills as the trigger of the end- 415 Permian mass extinction, Nat. Commun., 8(1), 164, doi:10.1038/s41467-017-00083-9, 2017.

416 Clapham, M. E. and Payne, J. L.: Acidification, anoxia, and extinction: A multiple logistic regression analysis of 417 extinction selectivity during the Middle and Late Permian, Geology, 39(11), 1059–1062, doi:10.1130/G32230.1, 418 2011.

419 Colbath, G. K. and Grenfell, H. R.: Review of biological affinities of Paleozoic acid-resistant, organic-walled 420 eukaryotic algal microfossils (including “acritarchs”), Rev. Palaeobot. Palyno., 86(3–4), 287–314, 421 doi:10.1016/0034-6667(94)00148-D, 1995.

422 Cui, Y. and Kump, L. R.: Global warming and the end-Permian extinction event: Proxy and modeling 423 perspectives, Earth-Science Reviews, 149, 5–22, doi:10.1016/j.earscirev.2014.04.007, 2015.

424 Devillers, R. and de Vernal, A.: Distribution of dinoflagellate cysts in surface sediments of the northern North 425 Atlantic in relation to nutrient content and productivity in surface waters, Mar. Geol., 166(1), 103–124, 426 doi:10.1016/S0025-3227(00)00007-4, 2000.

427 Ellegaard, M.: Variations in dinoflagellate cyst morphology under conditions of changing salinity during the last 428 2000 years in the Limfjord, Denmark, Rev. Palaeobot. Palyno., 109(1), 65–81, doi:10.1016/S0034- 429 6667(99)00045-7, 2000.

430 Ellegaard, M., Lewis, J. and Harding, I.: Cyst–theca relationship, life cycle, and effects of temperature and 431 salinity on the cyst morphology of Gonyaulax Baltica Sp. Nov. (dinophyceae) from the Baltic Sea Area, J. 432 Phycol., 38(4), 775–789, doi:10.1046/j.1529-8817.2002.01062.x, 2002.

433 Evitt, W. R.: A discussion and proposals concerning fossil dinoflagellates, hystrichospheres, and acritarchs, I, 434 PNAS, 49(2), 158–164, 1963.

435 Fischer, A. G.: Brackish oceans as the cause of the Permo-Triassic marine faunal crisis, in Problems in 436 Palaeoclimatology, edited by A. E. M. Nairn, pp. 566–574, Interscience, London., 1964.

437 Grard, A., François, L. M., Dessert, C., Dupré, B. and Goddéris, Y.: Basaltic volcanism and mass extinction at 438 the Permo-Triassic boundary: Environmental impact and modeling of the global carbon cycle, Earth and 439 Planetary Science Letters, 234(1), 207–221, doi:10.1016/j.epsl.2005.02.027, 2005.

440 Grasby, S. E. and Beauchamp, B.: Latest Permian to Early Triassic basin-to-shelf anoxia in the Sverdrup Basin, 441 Arctic Canada, Chem. Geol., 264(1–4), 232–246, doi:10.1016/j.chemgeo.2009.03.009, 2009.

442 Grasby, S. E., Beauchamp, B., Bond, D. P. G., Wignall, P., Talavera, C., Galloway, J. M., Piepjohn, K., 443 Reinhardt, L. and Blomeier, D.: Progressive environmental deterioration in northwestern Pangea leading to the 444 latest Permian extinction, Geol. Soc. Am. Bull., B31197.1, doi:10.1130/B31197.1, 2015.

445 Hallam, A. and Cohen, J. M.: The Case for Sea-Level Change as a Dominant Causal Factor in Mass Extinction 446 of Marine Invertebrates [and Discussion], Philos. T. Roy. Soc. B, 325(1228), 437–455, doi:10.2307/2396934, 447 1989.

448 Hammen, C. S. and Lum, S. C.: Salinity tolerance and pedicle regeneration of Lingula, J. Paleontol., 51(3), 548– 449 551, 1977.

450 Hotinski, R. M., Bice, K. L., Kump, L. R., Najjar, R. G. and Arthur, M. A.: Ocean stagnation and end-Permian 451 anoxia, Geology, 29(1), 7–10, doi:10.1130/0091-7613(2001)029<0007:OSAEPA>2.0.CO;2, 2001.

452 Isozaki, Y.: Permo-Triassic Boundary Superanoxia and Stratified Superocean: Records from Lost Deep Sea, 453 Science, 276(5310), 235–238, doi:10.1126/science.276.5310.235, 1997.

454 Jansonius, J.: Palynology of permian and triassic sediments, Peace River area, Western Canada, Palaeontogr. 455 Abt. B, 35–98, 1962.

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

456 Joachimski, M. M., Lai, X., Shen, S., Jiang, H., Luo, G., Chen, B., Chen, J. and Sun, Y.: Climate warming in the 457 latest Permian and the Permian–Triassic mass extinction, Geology, 40(3), 195–198, doi:10.1130/G32707.1, 458 2012.

459 Kearsey, T., Twitchett, R. J., Price, G. D. and Grimes, S. T.: Isotope excursions and palaeotemperature estimates 460 from the Permian/Triassic boundary in the Southern Alps (Italy), Palaeogeogr. Palaeocl., 279(1–2), 29–40, 461 doi:10.1016/j.palaeo.2009.04.015, 2009.

462 Korte, C. and Kozur, H. W.: Carbon-isotope stratigraphy across the Permian–Triassic boundary: A review, J. 463 Asian Earth Sci., 39(4), 215–235, doi:10.1016/j.jseaes.2010.01.005, 2010.

464 Lei, Y., Servais, T., Feng, Q. and He, W.: The spatial (nearshore–offshore) distribution of latest Permian 465 phytoplankton from the Yangtze Block, South China, Palaeogeogr. Palaeocl., 363–364, 151–162, 466 doi:10.1016/j.palaeo.2012.09.010, 2012.

467 Lei, Y., Servais, T., Feng, Q. and He, W.: Latest Permian acritarchs from South China and the 468 Micrhystridium/Veryhachium complex revisited, Palynology, 37(2), 325–344, 469 doi:10.1080/01916122.2013.793625, 2013a.

470 Lei, Y., Servais, T. and Feng, Q.: The diversity of the Permian phytoplankton, Rev. Palaeobot. Palyno., 198, 471 145–161, doi:10.1016/j.revpalbo.2013.03.004, 2013b.

472 Li, J., Servais, T., Yan, K. and Zhu, H.: A nearshore–offshore trend in acritarch distribution from the Early– 473 Middle Ordovician of the Yangtze Platform, South China, Rev. Palaeobot. Palyno., 130(1–4), 141–161, 474 doi:10.1016/j.revpalbo.2003.12.005, 2004.

475 Looy, C. V., Twitchett, R. J., Dilcher, D. L., Van Konijnenburg-Van Cittert, J. H. and Visscher, H.: Life in the 476 end-Permian dead zone, PNAS, 98(14), 7879–7883, 2001.

477 McGhee, G. R., Sheehan, P. M., Bottjer, D. J. and Droser, M. L.: Ecological ranking of Phanerozoic biodiversity 478 crises: The Serpukhovian (early Carboniferous) crisis had a greater ecological impact than the end-Ordovician, 479 Geology, 40(2), 147–150, doi:10.1130/G32679.1, 2012.

480 Mertens, K. N., Ribeiro, S., Bouimetarhan, I., Caner, H., Combourieu Nebout, N., Dale, B., De Vernal, A., 481 Ellegaard, M., Filipova, M., Godhe, A., Goubert, E., Grøsfjeld, K., Holzwarth, U., Kotthoff, U., Leroy, S. A. G., 482 Londeix, L., Marret, F., Matsuoka, K., Mudie, P. J., Naudts, L., Peña-Manjarrez, J. L., Persson, A., Popescu, S.- 483 M., Pospelova, V., Sangiorgi, F., van der Meer, M. T. J., Vink, A., Zonneveld, K. A. F., Vercauteren, D., 484 Vlassenbroeck, J. and Louwye, S.: Process length variation in cysts of a dinoflagellate, Lingulodinium 485 machaerophorum, in surface sediments: Investigating its potential as salinity proxy, Mar. Micropaleontol., 486 70(1–2), 54–69, doi:10.1016/j.marmicro.2008.10.004, 2009.

487 Mertens, K. N., Dale, B., Ellegaard, M., Jansson, I.-M., Godhe, A., Kremp, A. and Louwye, S.: Process length 488 variation in cysts of the dinoflagellate Protoceratium reticulatum, from surface sediments of the Baltic– 489 Kattegat–Skagerrak estuarine system: a regional salinity proxy, Boreas, 40(2), 242–255, doi:10.1111/j.1502- 490 3885.2010.00193.x, 2011.

491 Mertens, K. N., Bringué, M., Van Nieuwenhove, N., Takano, Y., Pospelova, V., Rochon, A., De Vernal, A., 492 Radi, T., Dale, B., Patterson, R. T., Weckström, K., Andrén, E., Louwye, S. and Matsuoka, K.: Process length 493 variation of the cyst of the dinoflagellate Protoceratium reticulatum in the North Pacific and Baltic-Skagerrak 494 region: calibration as an annual density proxy and first evidence of pseudo-cryptic speciation, J. Quaternary Sci., 495 27(7), 734–744, doi:10.1002/jqs.2564, 2012.

496 Mettam, C., Zerkle, A. L., Claire, M. W., Izon, G., Junium, C. J. and Twitchett, R. J.: High-frequency 497 fluctuations in redox conditions during the latest Permian mass extinction, Palaeogeogr. Palaeocl., 498 doi:10.1016/j.palaeo.2017.06.014, 2017.

499 Mudie, P. J., Aksu, A. E. and Yasar, D.: Late Quaternary dinoflagellate cysts from the Black, Marmara and 500 Aegean seas: variations in assemblages, morphology and paleosalinity, Mar. Micropaleontol., 43(1–2), 155–178, 501 doi:10.1016/S0377-8398(01)00006-8, 2001.

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

502 Mudie, P. J., Rochon, A., Aksu, A. E. and Gillespie, H.: Dinoflagellate cysts, freshwater algae and fungal spores 503 as salinity indicators in Late Quaternary cores from Marmara and Black seas, Mar. Geol., 190(1), 203–231, 504 doi:10.1016/S0025-3227(02)00348-1, 2002.

505 Pacton, M., Gorin, G. E. and Vasconcelos, C.: Amorphous organic matter — Experimental data on formation 506 and the role of microbes, Rev. Palaeobot. Palyno., 166(3–4), 253–267, doi:10.1016/j.revpalbo.2011.05.011, 507 2011.

508 Parrish, J. T., Ziegler, A. M. and Scotese, C. R.: Rainfall patterns and the distribution of coals and evaporites in 509 the Mesozoic and Cenozoic, Palaeogeogr. Palaeocl., 40(1), 67–101, doi:10.1016/0031-0182(82)90085-2, 1982.

510 Peng, Y., Shi, G. R., Gao, Y., He, W. and Shen, S.: How and why did the Lingulidae (Brachiopoda) not only 511 survive the end-Permian mass extinction but also thrive in its aftermath?, Palaeogeogr. Palaeocl., 252(1–2), 512 118–131, doi:10.1016/j.palaeo.2006.11.039, 2007.

513 Piasecki, S.: Preliminary palynostratigraphy of the Permian-Lower Triassic sediments in Jameson Land and 514 Scoresby Land, East Greenland, B. Geol. Soc. Denmark, 32, 139–144, 1984.

515 Pospelova, V., Chmura, G. L. and Walker, H. A.: Environmental factors influencing the spatial distribution of 516 dinoflagellate cyst assemblages in shallow lagoons of southern New England (USA), Rev. Palaeobot. Palyno., 517 128(1–2), 7–34, doi:10.1016/S0034-6667(03)00110-6, 2004.

518 Rodland, D. L. and Bottjer, D. J.: Biotic Recovery from the End-Permian Mass Extinction: Behavior of the 519 Inarticulate Brachiopod Lingula as a Disaster Taxon, PALAIOS, 16(1), 95–101, doi:10.1669/0883- 520 1351(2001)016<0095:BRFTEP>2.0.CO;2, 2001.

521 Sarjeant, W. A. S.: Acritarchs and tasmanitids form the Chhidru Formation, uppermost Permian of West 522 Pakistan, Stratigraphic Boundary Problems: Permian and Triassic of West Pakistan, Spec. Publ. 4, 277–304, 523 1970.

524 Sarjeant, W. A. S. and Stancliffe, R. P. W.: The Micrhystridium and Veryhachium Complexes (Acritarcha: 525 Acanthomorphitae and Polygonomorphitae): A Taxonomic Reconsideration, Micropaleontology, 40(1), 1–77, 526 doi:10.2307/1485800, 1994.

527 Schobben, M., Joachimski, M. M., Korn, D., Leda, L. and Korte, C.: Palaeotethys seawater temperature rise and 528 an intensified hydrological cycle following the end-Permian mass extinction, Gondwana Res., 26(2), 675–683, 529 doi:10.1016/j.gr.2013.07.019, 2014.

530 Sephton, M. A., Looy, C. V., Brinkhuis, H., Wignall, P. B., De Leeuw, J. W. and Visscher, H.: Catastrophic soil 531 erosion during the end-Permian biotic crisis, Geology, 33(12), 941–944, 2005.

532 Servais, T., Stricanne, L., Montenari, M. and Pross, J.: Population dynamics of galeate acritarchs at the 533 Cambrian–Ordovician transition in the Algerian Sahara, Palaeontology, 47(2), 395–414, doi:10.1111/j.0031- 534 0239.2004.00367.x, 2004.

535 Shen, J., Lei, Y., Algeo, T. J., Feng, Q., Servais, T., Yu, J. and Zhou, L.: Volcanic effects on microplankton 536 during the Permian-Triassic transition (Shangsi and Xinmin, South China), PALAIOS, 28(8), 552–567, 537 doi:10.2110/palo.2013.p13-014r, 2013.

538 Sluijs, A. and Brinkhuis, H.: A dynamic climate and ecosystem state during the Paleocene-Eocene Thermal 539 Maximum: inferences from dinoflagellate cyst assemblages on the New Jersey Shelf, Biogeosciences, 6(8), 540 1755–1781, doi:10.5194/bg-6-1755-2009, 2009.

541 Stemmerik, L. A. R. S., Bendix-Almgreen, S. E. and Piasecki, S.: The Permian–Triassic boundary in central 542 East Greenland: past and present views." Bulletin of the Geological Society of Denmark, B. Geol. Soc. 543 Denmark, 48(2), 159–167, 2001.

544 Stevens, C. H.: Was development of brackish oceans a factor in Permian extinctions?, Geol. Soc. Am. Bull., 545 88(1), 133–138, doi:10.1130/0016-7606(1977)88<133:WDOBOA>2.0.CO;2, 1977.

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

546 Stricanne, L., Munnecke, A., Pross, J. and Servais, T.: Acritarch distribution along an inshore–offshore transect 547 in the Gorstian (lower Ludlow) of Gotland, Sweden, Rev. Palaeobot. Palyno., 130(1–4), 195–216, 548 doi:10.1016/j.revpalbo.2003.12.007, 2004.

549 Sun, Y., Joachimski, M. M., Wignall, P. B., Yan, C., Chen, Y., Jiang, H., Wang, L. and Lai, X.: Lethally hot 550 temperatures during the Early Triassic greenhouse, Science, 338(6105), 366–370, doi:10.1126/science.1224126, 551 2012.

552 Surlyk, F., Piasecki, S., Rolle, F., Stemmerik, L., Thomsen, E. and Wrang, P.: The Permian base of East 553 Greenland, in Petroleum Geology of the North European Margin, edited by A. M. Spencer, pp. 303–315, 554 Springer Netherlands., 1984.

555 Svensen, H., Planke, S., Polozov, A. G., Schmidbauer, N., Corfu, F., Podladchikov, Y. Y. and Jamtveit, B.: 556 Siberian gas venting and the end-Permian environmental crisis, Earth Planet. Sc. Lett., 277(3–4), 490–500, 557 doi:10.1016/j.epsl.2008.11.015, 2009.

558 Tappan, H. and Loeblich, A. R.: Evolution of the oceanic plankton, Earth-Sci. Rev., 9(3), 207–240, 559 doi:10.1016/0012-8252(73)90092-5, 1973.

560 Twitchett, R. J.: The Lilliput effect in the aftermath of the end-Permian extinction event, Palaeogeogr. Palaeocl., 561 252(1–2), 132–144, doi:10.1016/j.palaeo.2006.11.038, 2007.

562 Twitchett, R. J., Looy, C. V., Morante, R., Visscher, H. and Wignall, P. B.: Rapid and synchronous collapse of 563 marine and terrestrial ecosystems during the end-Permian biotic crisis, Geology, 29(4), 351–354, 564 doi:10.1130/0091-7613(2001)029<0351:RASCOM>2.0.CO;2, 2001.

565 Tyson, R.: Sedimentary Organic Matter: Organic facies and palynofacies, 1995 edition., Springer., 1995.

566 Vernal, A. de, Hillaire-Marcel, C., Turon, J.-L. and Matthiessen, J.: Reconstruction of sea-surface temperature, 567 salinity, and sea-ice cover in the northern North Atlantic during the last glacial maximum based on dinocyst 568 assemblages, Can. J. Earth Sci., 37(5), 725–750, doi:10.1139/e99-091, 2000.

569 Westerbom, M., Kilpi, M. and Mustonen, O.: Blue mussels, Mytilus edulis, at the edge of the range: population 570 structure, growth and biomass along a salinity gradient in the north-eastern Baltic Sea, Mar. Biol., 140(5), 991– 571 999, doi:10.1007/s00227-001-0765-6, 2002.

572 Wignall, P. B. and Hallam, A.: Anoxia as a cause of the Permian/Triassic mass extinction: facies evidence from 573 northern Italy and the western United States, Palaeogeogr. Palaeocl., 93(1), 21–46, doi:10.1016/0031- 574 0182(92)90182-5, 1992.

575 Wignall, P. B. and Twitchett, R. J.: Oceanic anoxia and the end Permian mass extinction, Science, 272(5265), 576 1155, 1996.

577 Wignall, P. B. and Twitchett, R. J.: Permian–Triassic sedimentology of Jameson Land, East Greenland: incised 578 submarine channels in an anoxic basin, J. Geol. Soc., 159(6), 691–703, doi:10.1144/0016-764900-120, 2002.

579 Winguth, A. and Winguth, C.: Precession-driven monsoon variability at the Permian–Triassic boundary — 580 Implications for anoxia and the mass extinction, Global. Planet. Change, 105, 160–170, 581 doi:10.1016/j.gloplacha.2012.06.006, 2012.

582 Yin, H., Ed.: The Palaeozoic-Mesozoic boundary, candidates of Global Stratotype Section and Point of the 583 Permian-Triassic boundary, China University of Geosciences Press, Wuhan., 1996.

584 Zonneveld, J.-P., Beatty, T. W. and Pemberton, S. G.: Lingulide Brachiopods and the Trace Fossil Lingulichnus 585 from the Triassic of Western Canada: Implications for Faunal Recovery After the End-Permian Mass Extinction, 586 PALAIOS, 22(1), 74–97, doi:10.2110/palo.2005.p05-103r, 2007.

587 Zonneveld, K. A. F., Marret, F., Versteegh, G. J. M., Bogus, K., Bonnet, S., Bouimetarhan, I., Crouch, E., de 588 Vernal, A., Elshanawany, R., Edwards, L., Esper, O., Forke, S., Grøsfjeld, K., Henry, M., Holzwarth, U., Kielt,

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

589 J.-F., Kim, S.-Y., Ladouceur, S., Ledu, D., Chen, L., Limoges, A., Londeix, L., Lu, S.-H., Mahmoud, M. S., 590 Marino, G., Matsouka, K., Matthiessen, J., Mildenhal, D. C., Mudie, P., Neil, H. L., Pospelova, V., Qi, Y., Radi, 591 T., Richerol, T., Rochon, A., Sangiorgi, F., Solignac, S., Turon, J.-L., Verleye, T., Wang, Y., Wang, Z. and 592 Young, M.: Atlas of modern dinoflagellate cyst distribution based on 2405 data points, Rev. Palaeobot. Palyno., 593 191, 1–197, doi:10.1016/j.revpalbo.2012.08.003, 2013.

594

595

596

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

597 Figure 1. Modern map of Jameson Land, East Greenland (adapted from Piasecki, 1984). The location of the 598 studied section is indicated with a red circle. The right panel show the lithology of the Jameson Land section 599 and sample depths. (* spore-peak after Looy et al., 2001, marine collapse after Twitchett et al., 2001).

600

601 602

603

604

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

605 Figure. 2. Change in major groups of palynofacies. AOM= amorphous organic matter. The grey shaded box

606 highlights the biotic crisis as defined by the high spore/pollen ratios, which are indicative for the disappearance

607 of forest communities. Green shading indicates data from bioturbated rocks; grey shading indicates data from

608 laminated rocks. Grey horizontal bar indicates the extinction interval, based on high spore: pollen ratios.

609

610

611

612

613

614

615

616

617

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

618 Figure 3. Aquatic palynomorphs plate. A) M. pentagonale – group, B) V. laidii - group C) Cymatiosphaera sp. 619 D) and E) leiospheres F)-I) M. breve -group J) M. pentagonale - group K)-M) M. breve - group N) 620 Cymatiosphaera sp. O) and P) M. breve -group

621 622

623

624

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

625 Figure 4. Aquatic palynomorphs. The figure showsconcentration of all aquatic palynomorphs andrelative

626 abundance specific acritarch and prasinophyte groups. Arrow indicate intervals with low counts (<20). Green

627 shading indicates data from bioturbated rocks; grey shading indicates data from laminated rocks. Grey

628 horizontal bar indicates the extinction interval, based on high spore: pollen ratios.

629

630

631

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

632 Figure 5. Size-frequency of body diamter (left) and proces length (right).

633

634

635

Clim. Past Discuss., https://doi.org/10.5194/cp-2017-136 Manuscript under review for journal Clim. Past Discussion started: 25 October 2017 c Author(s) 2017. CC BY 4.0 License.

636 Figure 6. Changes in acritarch abundance and morphology. In the panel on the left are the main groups of

637 aquatic palynomorphs summarized. On the right: body size, process length, and process length relative to body

638 size. Where are low counts? Green shading indicates data from bioturbated rocks; grey shading indicates data

639 from laminated rocks. The grey bar indicates the extinction interval, based on high spore:pollen ratios. The five

640 small black arrows indicate in which samples the number of specimens used for measurements was low (4-16

641 specimens). Grey horizontal bar indicates the extinction interval, based on high spore: pollen ratios.

642

643

644